Why learning redox is difficult even with animations – students’ adherence to cognitive resources

Abstract

Students face many challenges when they learn redox reactions at the submicroscopic level. Animations have been shown to support students’ learning, but some students may not interpret the animations as intended. Guided by the cognitive resources, or the Knowledge in Pieces perspective, this study examined how students made sense of redox reactions after they engaged in different animation activities. Our participants were 15 first-year university students taking introductory chemistry courses. Besides their written responses to diagram-based post-animation questions, we also interviewed six of these 15 participants so that they could elaborate on their written responses. We identified two cognitive resources to which students strongly adhered when reasoning about redox reactions. The first, ion pair, reflects the idea that ions remain bonded as pairs in solution. The second, ion attraction, frames redox reactions as interactions between oppositely charged ions. While these ideas have been reported elsewhere, we argue that they are key to students’ sense-making of redox because, in determining the best representation of redox reaction, students primarily focused on whether ions appeared paired or whether opposite charges could “react”. Furthermore, we suggest that these two cognitive resources could explain students’ alternative conceptions in other chemistry topics such as bonding, structure and properties, solution chemistry, and elements/compounds. Our findings suggest that instruction and animation design could help students visualise ion distribution before electron transfer, and support students to see the values and limitations of ion pair and ion attraction in different contexts.

---

Introduction

Redox reactions are one of the most important topics in chemistry (De Jong and Treagust, 2003). They are involved in batteries for pacemakers, mobile phones, electric vehicles, and the storage of renewable energy sources. Many of the biochemical processes, such as photosynthesis and respiration, are redox reactions that involve the transfer of electrons across cascading systems of biomolecules. An understanding of redox reactions is thus essential for students to make sense of extensive chemical, physical, biological, and technological processes and applications.

Prior research in chemistry education has established that it is challenging to develop a scientific understanding of redox reactions (Österlund et al., 2010; Hadinugrahaningsih et al., 2022). Part of the challenge relates to the inherent nature of chemistry learning, which involves integrating macroscopic and submicroscopic phenomena as well as symbolic representations (Gilbert and Treagust, 2009; Taber, 2013; Sevian and Talanquer, 2014). At the submicroscopic level, students are expected to identify reacting and non-reacting species, mentally visualise the electron transfer between reacting species but not the spectator species, and identify oxidants and reductants accordingly (Brandriet and Bretz, 2014). Researchers and practitioners have been using molecular animations to make submicroscopic events and their linkages to the macroscopic and symbolic world more accessible to students (Tasker and Dalton, 2006; Kelly, 2016). Yet even these interventions do not always serve their intended purposes (Rosenthal and Sanger, 2012; Kelly, 2017; Akaygun and Adadan, 2019) or may still present challenges at some points, even when scaffolded (Magnone and Yezierski, 2024a, 2024b), partly because students can interpret the animations in different ways. This led us to ask, “What do students (mis)understand about representations of redox reactions?”

However, rather than focusing on this question, the current study draws on the cognitive resources, or Knowledge in Pieces, perspective (diSessa, 1993; Hammer, 1996) as an alternative way to think about this issue. According to this theoretical approach, student thinking is not governed by singular, stable conceptions but is dynamically constructed from multiple small-scale knowledge elements. These elements may be intuitive or formal, implicit or explicit; they are activated, reorganised, or refocused depending on the task and the context (Smith et al., 1994; Potvin and Cyr, 2017), and they are important cognitive resources for learning. Moreover, during learning, students may construct cognitive resources from multiple knowledge elements (Hammer, 2000). Rather than focusing on identifying and reporting full-blown “misconceptions,” this study examines these cognitive resources and how they may relate to each other in ways that inform students’ reasoning about submicroscopic structures and processes in redox reactions (Taber and García-Franco, 2010).

This understanding of cognition provides a useful lens for analysing how students engage with complex representations in chemistry, including redox animations at the submicroscopic level. To interpret these representations, students are expected to coordinate a range of cognitive resources that contribute to their understanding of phenomena such as electron transfer, the distribution of reacting and spectator ions in solutions, symbolic redox notation, and links between observable and particulate-level changes. This concept of cognitive resources has been used to advance our understanding of students’ learning in chemistry in some topics (Taber and García-Franco, 2010; Kelly et al., 2021; Pölloth et al., 2023; Kiernan et al., 2024). To date, no study has examined how students coordinate or assemble their cognitive resources when learning about redox at the submicroscopic level. The current study fills this gap.

Extending the cognitive resources perspective, Potvin and Cyr (2017) suggest that students often possess multiple conceptual resources, which may be reorganised or refocused depending on the context. Students’ responses to different tasks in different contexts are informed by whatever cognitive resources they adhere to most strongly and which of these play a dominant role in shaping their explanations (Potvin, 2017). With this in mind, the current study employed the framework of adherence and prevalence (Potvin and Cyr, 2017) to examine how students prioritised certain cognitive resources over others. While multiple resources may be activated, some gain stronger weight (adherence) and dominate reasoning in different contexts (prevalence). This lens allowed the study to focus on possible patterns of conceptual reasoning. Rather than asking “What do students (mis)understand about representations of redox reactions?”, we instead asked the following research question:

“What cognitive resources may students adhere to and coordinate in their efforts to interpret representations of redox reactions?”

This study is part of a broader investigation into how best to support student learning with animations; specifically, it compares the impact of viewing a scientifically accurate animation with the impact of critiquing multiple animations of varying accuracy. While the larger project would explore the effectiveness of these two strategies, there remained a lack of in-depth knowledge about how students interpret animations. Drawing upon these findings, the present study contributes to the design of animations and scaffolds that increase the likelihood of students using cognitive resources that are conducive to their learning.

A review of students’ understanding of redox reactions: existing research and knowledge gaps

Research has extensively examined students’ understanding of redox reactions, particularly their conceptual difficulties and misconceptions. For example, some students believe that the cations and anions of electrolytes remain bonded as ion pairs in aqueous solutions (Rosenthal and Sanger, 2012). And some also view electron transfer as a process that occurs when the bond between a cation and a spectator ion breaks or forms (Brandriet and Bretz, 2014). Other common difficulties include confusion over oxidation numbers and ionic charges, as well as identifying the correct oxidising and reducing agents (Brandriet and Bretz, 2014; Masykuri et al., 2019). More recent studies also show that students can recognise oxidation and reduction processes but struggle to understand electron transfer (Alqadri and Munawwarah, 2025).

Although these difficulties are often attributed to students’ inability to visualise the invisible submicroscopic aspects of redox reactions, it is noteworthy that such challenges persist even after students view animations that explicitly show redox processes at the molecular level. For instance, using Tasker and Dalton's (2006) widely adopted animation representing the redox reaction between solid copper and aqueous silver nitrate, Rosenthal and Sanger (2012) found that students continued to believe that cations and anions remain bonded as ion pairs in solution; they also misattributed nitrate ions (NO₃⁻) as the driving force of the redox reaction and misunderstood how electron transfer affects charge. Using a more interactive approach, Kelly (2017) asked students to critique two contrasting animations with different levels of chemical accuracy. While students were able to recognise the scientifically accurate animation (which illustrated electron transfer), nearly all of them perceived both animations as correct and useful. Many made errors when evaluating how each animation was supported or refuted by the experimental evidence and continued to rely on and reproduce features from the inaccurate version in their own representations. The findings reflect that students may have known which animation was accurate, but they still relied on intuitive or familiar ideas when constructing their own understanding. Similarly, Akaygun and Adadan (2019) found that, when students were asked to differentiate and critique scientifically accurate and inaccurate animations, most could correctly identify the accurate version. But only a small number explained electron transfer from zinc to copper ions, copper deposition on zinc, and the formation of zinc ions. Although the task of critiquing helped students recognise features such as electron transfer, the study highlights students’ persistent challenges when attempting to understand redox reactions. These studies suggest that the animations did not ensure students developed a comprehensive understanding of simple redox reactions. In other words, these animations only seemed to help some of the students to recognise some of the features of redox some of the time.

These persistent difficulties with students’ understanding highlight a critical issue: knowing that students hold incorrect ideas and identifying what those ideas are does not necessarily explain why students continue to rely on them. The previously mentioned studies might have assumed that students held these conceptions as relatively stable “stored constructs” (Hammer, 1996, p. 102), rather than constructing them on the spot during the study. An alternative way to conceptualise learning is that scientific explanations are not retrieved whole from memory but involve the person assembling quite a few smaller pieces of knowledge—cognitive resources—in real time based on the demands of the task (Hammer et al., 2005; Sherin et al., 2012). Analysing how students assemble these resources would provide a more nuanced understanding of conceptual reasoning. It may also help us understand if there are cognitive resources that students tend to “stick with” during their construction of concepts related to redox reactions. The study, therefore, explores students’ learning challenges by identifying cognitive resources that could be generic across contexts. We are particularly interested in how different conceptual resources might contribute to students’ understanding of animations that represent redox, as well as their comprehension of redox reactions more generally. For example, to what extent does it matter whether students think that the ions of electrolytes associate or dissociate in water? And if they indicate redox is a simple spatial rearrangement of particles, how might they reconcile animations that show electron transfer?

Theoretical framework: adherence and prevalence of cognitive resources in students’ learning

The notion of cognitive resources being assembled in real time, as discussed in the previous section, is grounded in diSessa's Knowledge in Pieces (KiP) framework, which views the structure of student knowledge not as a coherent, stable framework but as a collection of smaller-scale, fine-grained, and loosely connected knowledge elements that are built from minimal structures called phenomenological primitives (p-prims) (diSessa, 1993). These elements may be intuitive or formal, implicit or explicit, and they are activated based on tasks and contexts (Smith et al., 1994). P-Prims (e.g., “more effort yields more result”) arise from prior experience and often seem sensible to students. Of interest to education researchers is how p-prims are used in different contexts as resources for learning (Hammer, 1996; diSessa, 2017).

Understandings of chemical phenomena, including redox, are often grounded in structural entities rather than in the abstraction of processes in physics (Taber and García-Franco, 2010). Thus, we focus in this study on cognitive resources that students draw upon to explain phenomena (Hammer, 2000). Cognitive resources could be of bigger grain size than p-prims; they are also depending on the individual's tasks and contexts, shaped through prior experience, and assembled dynamically based on what feels most relevant at the time in response to specific tasks, prompts, or representations (Hammer, 2000; Sherin et al., 2012; diSessa, 2017).

Studies in chemistry education have demonstrated that learners used cognitive resources in different ways to explain chemical phenomena such as energetic and structural changes in chemical reactions (Pölloth et al., 2023), molecular geometry (Kiernan et al., 2024), kinetic behaviours of molecules (Rodriguez et al., 2020), and substitution reactions in organic chemistry (Zaimi et al., 2025). In the context of interpreting submicroscopic representations, Kelly et al. (2021) explored resources that students use to make sense of the atomic-level processes involved in a precipitation reaction. By using card sorting and modelling exercises, students were guided to compare animations to the models they constructed and to judge the accuracy of the animation components. The study found that students shifted between their resources and sometimes misapplied them. The resources were connected to the students’ understanding of ion dissociation, states of matter, electrical conductivity, and formula construction, as well as charge attraction/repulsion.

These studies illustrate how KiP foregrounds students’ dynamic activation of multiple resources when they are attempting to explain chemical phenomena. However, while KiP accounts for the variability of students’ ideas, it provides less explanation for what happens when resources compete, that is, whether one idea dominates, coexists, or is only activated situationally. In the context of redox, for example, while Rosenthal and Sanger (2012) and Brandriet and Bretz (2014) identified the ion pair as a conception commonly used by students to make sense of redox reactions, Teichert et al. (2008) found that students were more likely to use the ion pair conception when they explained dissolution than when they explained electrical conductivity. We suggest that this finding is best explained by the idea that students simultaneously hold multiple cognitive resources (coexistence), with some of them becoming more dominant due to their stronger adherence or greater prevalence in particular contexts (Potvin and Cyr, 2017). In this framework, adherence refers to how strongly a student holds onto a particular idea or cognitive resource, while prevalence describes which ideas or cognitive resources emerge more often during different reasoning tasks and/or in different contexts.

The concepts of adherence and prevalence have informed research into students’ reasoning in the context of science versus intuition (Shtulman and Young, 2024) and competitions versus cooperative behaviours in nature. These concepts have also been studied in light of the association between students’ views and their understanding of evolution (Shtulman, 2025), as well as students’ commitment to the notions that “moving things are alive” (Skelling-Desmeules et al., 2021). Adherence and prevalence have also been found useful to make sense of students’ learning of socio-scientific issues (Leung and Cheng, 2020) and critical thinking (Cheng and Leung, 2021). In all these studies, the researchers focused on examining participants’ shifts between two alternatives in different contexts. However, in the current study, which considered students’ learning of redox, students were asked to assemble a large number of cognitive resources, rather than focusing on picking between two, in order to make sense of electrochemical phenomena. Bringing in the concept of adherence to this process provides a layered account of students’ reasoning: KiP highlights the diversity and context sensitivity of the cognitive resources students activate, while adherence foregrounds the relative strength of particular cognitive resources to guide their explanations and their rejection of alternatives. To our knowledge, this study is the first to integrate KiP with adherence and prevalence in the context of redox and submicroscopic animations. It therefore offers novel insights into students’ reasoning across tasks. Fig. 1 represents some key aspects of our framework.

Fig. 1 An illustration of conceptual adherence across students. Each shape represents a cognitive resource, with those in bold representing cognitive resources that are activated. Both students persistently use the ‘pentagon’ in different prompts and questions, meaning that this cognitive resource has a strong adherence.

Methods

(1) Ethical considerations and participants

This study received ethical approval from the University of Auckland Human Participants Ethics Committee (Ref: UAHPEC28338) on 14 October 2024. Participants were informed about the study through in-class announcements, Canvas posts, and a Participant Information Sheet, and they gave consent voluntarily via a Qualtrics online form. The researcher was not part of the teaching team, and lecturers were not told who did or did not participate. Students were assured that participation would not affect their course grades. All students had access to the learning materials, and a cross-treatment design ensured no group was disadvantaged. Data were anonymised and accessible only to the research team.

(2) Students’ engagement with the animations and post-activity questions

As previously mentioned, the study reported in this paper contributes to a larger research project that investigates how different uses of animation can support students’ conceptual understandings in chemistry. Specifically, compares two instructional approaches: (1) viewing a single scientifically accurate animation and (2) critiquing three animations with varying scientific accuracy. This paper does not evaluate which approach is more effective; rather, it focuses on how students make sense of redox reactions at the submicroscopic level, particularly, which conceptual resources they adhere to and how those resources shape their interpretation of animations. Nevertheless, we acknowledge that these instructional experiences, whether viewing or critiquing animations, may have influenced which conceptual resources were activated or adhered to during students’ reasoning.

After we received ethics committee approval, participants were recruited from two undergraduate chemistry courses via a voluntary sign-up link posted on their online learning platforms. Participants were first-year undergraduates enrolled in two introductory chemistry courses. Before participating, they had been taught redox at the macroscopic and symbolic levels (oxidation–reduction principles, oxidising and reducing agents, spectator ions, oxidation numbers, and half-reactions), and they had completed laboratory experiments focusing on observable changes. However, they had not yet encountered submicroscopic representations of redox, which became the focus of the animation design.

Participants were randomly assigned to either the viewing group or the critiquing group. They engaged with animations that represented a redox reaction between copper metal and silver nitrate solution (Fig. 2). Details of the animation tasks and the animations are provided in the SI.

Fig. 2 Screenshots of the copper–silver nitrate animation (viewed by students before doing the post-activity questions).

Following the animation tasks, students completed diagram-based post-activity questions, and their responses provided data about their understanding of redox. The questions were based on two reactions (shown by experiment videos) that were not used in the animations: (1) chlorine solution (Cl₂) with potassium iodide (KI) solution, and (2) zinc (Zn) reacting with copper(II) sulphate solution (CuSO₄).

After watching a video of the experiment for each reaction, the students were prompted to identify the best submicroscopic representation of the corresponding reaction out of three options. Fig. 3 presents the options for these two reactions. The inaccurate aspects reflected common alternative conceptions that have been identified in prior studies (Rosenthal and Sanger, 2012; Brandriet and Bretz, 2014; Hadinugrahaningsih et al., 2022), which varied with regard to (1) ion distribution, (2) electron transfer mechanisms, and (3) charge representation. Explanations about the accuracy of each option included in the diagram-based questions are presented in the SI.

Fig. 3 Post-activity questions.

In the post-activity questions, students were presented with multiple diagrams of two redox reactions at the submicroscopic level that varied in terms of (1) how ions were distributed, (2) which species were involved in electron transfer, and (3) how charge was represented. We asked the following five questions for each reaction:

(1) Which diagram best represents the reaction?

(2) Why is that diagram the best?

(3) Why are the other diagrams inaccurate?

(4) How does the macroscopic change (e.g., colour shift) relate to submicroscopic processes?

(5) Which species underwent oxidation and reduction?

These questions were designed to elicit not only the kind of ideas students activated but also which ideas they trusted, used, and reused to justify or reject competing representations among the multiple ideas that were activated. In other words, the questions attempted to identify students’ adherence to certain cognitive resources. The full procedure for data collection is shown in Fig. 4.

Fig. 4 Procedural flowchart.

Interviews. Individual semi-structured interviews were conducted in English with three students from each group to gain a deeper understanding of the students’ reasoning for both the animation activities and their responses to the post-activity questions. The first author acted as the interviewer. Students were shown the activities and their previous responses and were then asked to explain their choices; for example, “Can you explain why you answered like this back then?” “What makes this diagram different from the one you rejected?” These questions attempted to determine what kind of cognitive resources were activated. Students were then asked follow-up questions that referred back to their earlier responses to the post-activity questions in order to better understand their adherence. Examples of these questions included “Why do you think ions should be together?” “Which species are actually involved in the electron transfer?” “Which species would attract each other here?” and “Does being +/− make them more likely to react?” To minimise the risk of misinterpretation, the interviewer often repeated students’ answers and asked for confirmation using prompts such as “Is that what you mean?” or “Am I right in understanding this as…?” The interviews, with each lasted for 20–30 minutes, were audio-recorded and transcribed for analysis.

Data analysis. Students’ data of the post-activity questions were analysed according to five aspects of the redox reactions (Table 1). These aspects were based on a previous redox study (Brandriet and Bretz, 2014). Scientific understanding of redox reasoning involves an integrated and coherent understanding of these aspects. Prior studies have shown that some students relied on several of these aspects in their efforts to explain redox reactions, and they had different ways of making sense of these aspects. Therefore, our analysis took one more step in looking into students’ responses regarding how ions were distributed (for “1” in Table 1), whether they involved electron transfer (for “2”), and which reacting species were involved (for “3”), etc. The analysis (or breaking down) of these aspects was deemed essential because students could indicate electron transfer and refer to a variety of species they believed were involved in the reactions. This approach reflects the KiP perspective, in which we were able to capture students’ different ways of making sense of redox and hence determined their cognitive resources and their adherence to those resources.

Table 1 Aspects of understanding the redox reactions in post-activity questions

Cognitive resources	Key indicators of student responses
Ion distribution
1. Ionic compounds (KI and CuSO4) dissociate into ions when dissolved in aqueous solution.	• Described K^+ and I^− or Cu^2+ and SO4^2− as separate ions in solution.
	• Accepted diagrams showing free-floating ions before the reaction.
Electron transfer
2. Redox reactions involve a transfer of electrons between species, rather than “switching partners”.	• Explained that electrons were transferred from I^− to Cl2 or Zn to Cu^2+.
	• Rejected explanations based on “switching partners.”
Reacting species
3. Redox reactions involve particular chemical species (i.e., I^− and Cl2 and Zn and Cu^2+), while others are spectator ions (e.g., K^+, SO4^2−).	• Recognised that I^− and Cl2 react (not K^+), or Zn and Cu^2+ react (not SO4^2−).
	• Mentioned spectator ions appropriately or ignored them.
Symbolic redox knowledge
4. There are oxidation and reduction in a redox reaction, or it involves an oxidising agent and a reducing agent.	• Stated changes like “I^− go from −1 to 0 = oxidation.”
	• Used correct redox terminology (oxidation, reduction, oxidant, reductant).
Linking macro to submicro
5. There is a connection between macroscopic observations (e.g., colour change) with submicroscopic events (e.g., electron transfer, species change).	• Explained that I2 caused brown colour, or Cu^2+ removal faded the blue colour.
	• Explained how visible change reflected particle-level changes.

---

In addition, we used the following three data sources from the post-activity questions to identify the cognitive resources students adhered to:

(i) the diagram they selected;

(ii) their written justification for selecting the diagram;

(iii) their written explanation and reasoning for rejecting the alternatives and their responses to questions (4) and (5) of the post-activity questions.

During the interviews, participants were also asked to elaborate orally on (ii) and (iii) points above. Through qualitative analysis, we identified adherence when some ideas were repeatedly mentioned across post-activity questions (1)–(3) and when students consistently chose and prioritised one idea over other available ideas that they also activated during their reasoning, even when those alternatives were scientifically correct and in contradiction. We then counted the total number of responses that were identified as adherence across different reactions, and when the cognitive resource appeared in different reactions, it was identified as being prevalent.

Previous studies have measured conceptual adherence mainly through behavioural or neurocognitive indicators, such as reaction times (Shtulman and Valcarcel, 2012), EEG signals (Skelling-Desmeules et al., 2021), or argumentative writing tasks (Leung and Cheng, 2020). In contrast, this study adopted a qualitative reasoning-based approach that identified adherence through students’ representational choices, explanations, and justifications. This approach builds on and adds value to earlier methods by demonstrating that adherence can also be identified qualitatively through consistent patterns in students’ reasoning and offers a new approach to measuring and understanding adherence in chemistry education.

All procedures were documented in detail to maintain dependability and provide a transparent record of analytic decisions. To enhance credibility, interpretations were reviewed and refined through iterative discussions between the authors alongside triangulation of written and interview data. Although a formal second-coder check was not conducted, coding and thematic interpretations were reviewed and refined between authors to enhance consistency. Findings were grounded in students’ actual responses and corroborated across data sources to support confirmability.

Results

A total of 33 students signed up for the study. Fifteen of them completed the questions, eight from the viewing group and seven from the critiquing group. Table 2 summarises the number of students in each group who chose the best representation for the two reactions from post-activity questions. The primary aim of this study was to identify cognitive resources that students may use and adhere to when coordinating their interpretation of redox reactions. Given the small number of participants, we could not draw any meaningful conclusions about the effectiveness of critiquing or viewing.

Table 2 The number of students in each group who chose the best submicro representation for the reactions

Reactions	Viewing group (n = 8)	Critiquing group (n = 7)
R1: KI + Cl2	4	1
R2: Zn + CuSO4	3	4

---

To answer our research question, we analysed students’ written responses and interview data relating to the reactions in the post-activity questions according to the five categories in Table 1. The analysis is summarised in Table 3. Students in the viewing group and the critiquing group are coded as A1–A7 and B1–B8, respectively.

Table 3 Students’ options for the two post-activity questions, and an analysis of their written responses and the interview data

Students & chosen option		Aspects of redox reactions
		Ion distribution	Electron transfer	Reacting species	Symbolic redox	Linking macro & submicro
Students marked with an asterisk (*) were interviewed. The best option for the post-activity question (Option A in R1, Option B in R2) is in bold text.
A1*	R1: A	Dissociated	Yes	I^− + Cl2	Yes	Yes
	R2: B	Dissociated	Yes	Cu^2+ + Zn	Yes	Yes
A2	R1: A	Dissociated	Yes	I^− + Cl2	Yes	Yes
	R2: B	Dissociated	Yes	Cu^2+ + Zn	Yes	Yes
A3	R1: A	Dissociated	Yes	I^− + Cl2	Yes	Yes
	R2: B	Dissociated	Yes	Cu^2+ + Zn	Yes	Yes
A4	R1: A	Dissociated	Yes	I^− + Cl2	Yes	Yes
	R2: A	—	Yes	Cu^2+ + Zn	Yes	Yes
A5*	R1: C	Ion pairs	Yes	K^+ + Cl2	Yes	Yes
	R2: A	Ion pairs	Yes	Cu^2+ + Zn	Yes	Yes
A6*	R1: C	Ion pairs	Yes	K^+ + Cl2	Yes	Yes
	R2: A	Ion pairs	Yes	Cu^2+ + Zn	Yes	Yes
A7	R1: C	Ion pairs	No	K^+ + Cl2	Yes	Yes
	R2: C	Ion pairs	No	Cu^2+ + Zn	Yes	Yes
A8	R1: B	Dissociated	No	I^− + Cl2	—	—
	R2: —	—	—	—	—	—
B1	R1: A	Dissociated	Yes	I^− + Cl2	Yes	Yes
	R2: B	Dissociated	Yes	Cu^2+ + Zn	Yes	Yes
B2*	R1: C	Ion pairs	Yes	K^+ + Cl2	Yes	Yes
	R2: B	Dissociated	Yes	Cu^2+ + Zn	Yes	Yes
B3	R1: B	—	Yes	K^+ + Cl2	Yes	No
	R2: B	Dissociated	Yes	Cu^2+ + Zn	Yes	No
B4	R1: B	—	Yes	I^− + Cl2	Yes	No
	R2: —	—	—	Cu^2+ + Zn	—	—
B5*	R1:C	Ion pairs	No	I^− + Cl2	Yes	No
	R2: C	Ion pairs	Yes	Cu^2+ + Zn	Yes	No
B6	R1:C	Ion pairs	No	I^− + Cl2	Yes	No
	R2: B	Dissociated	No	Cu^2+ + Zn	Yes	No
B7*	R1: C	Ion pairs	—	I^−	—	—
	R2: C	Ion pairs	Yes	—	—	Yes

---

The details of the analysis in Table 3 are as follows:

a. Ion distribution (as “dissociated” or “ion pairs”)

b. Electron transfer (as “Yes” or “No,”) depending on whether they mentioned gaining/losing or transfer of electrons

c. Reacting species (as “I⁻/Cl₂” or “K” in Reaction 1; as “Cu²⁺/Zn²⁺” in reaction 2)

d. Symbolic redox (as “Yes” when students identified the oxidising agent and reducing agent, identified which chemical species underwent oxidation and reduction, or stated the oxidation numbers of the reacting species; and “No” otherwise)

e. Linking macro (as “Yes” when the responses referred to observable changes with respect to changes in chemical species, and “No” otherwise)

f. When students did not provide any written responses related to particular aspects or did not refer to them during the interview, we use the “—” sign in the table.

The table is not to be read as a scoring table like the type we use in summative assessment. Instead, it summarises whether students referred to particular aspects of the reactions. Take “electron transfer” as an example: as long as students mentioned it in their responses, we would mark that as “Yes,” irrespective of the reacting species. In the “reacting species” column, we recorded what reacting species students mentioned, and we indicated when students identified which chemical species underwent oxidation and reduction in the “symbolic” column.

From the data, we identified two underlying cognitive resources that students seemed to have adhered to, which guided their reasoning about redox reactions while answering the post-activity questions. Table 4 provides a summary of how we constructed these two cognitive resources based on a range of data sources and observations. These two cognitive resources are reported in the following subsections.

Table 4 Observed features showing strong adherence

Cognitive resources	Observed features showing strong adherence
Ion-pair	• Chosen diagram showing bonded ions (e.g. KI as a pair, CuSO4 as a pair).
	• Justifications: Repeatedly stated ions “must stay together”, “are a compound”.
	• Rejection of alternatives: Eliminated other diagrams because they showed dissociation “should not separate”.
	• Others: Referred to electron transfer or redox agents only when asked directly, but used ion-pairs to determine the best diagram.
Ion-attraction	• Chosen diagram: Selected diagrams that showed positive species reacting with negative species (e.g. K^+ with Cl^−) even when this was not the correct reacting pair.
	• Justifications: Statements like “positive attracts negative”.
	• Rejection: Eliminated correct options because the reacting species were not opposite charges. “I^− and Cl will not react because same charge”, or “I and Cl are halogens so they cannot react”.
	• Other: Sometimes paired with electron transfer but still explained reactions in terms of attraction forces.

---

(1) Ion-pair cognitive resource

The key visual cues for the choices in the post-activity questions were (1) whether the ions were dissociated or not, and (2) whether there was electron transfer. Among the students who selected the inaccurate options (B or C in R1 and A or C in R2), the most common justification was that KI (in R1) and CuSO₄ (in R2) should remain bonded together in the solutions; that is, these compounds exist as ion pairs. Also, some students (A5, A6, A7, B5, and B7) rejected the best option, which represented electron transfer, because the electrolytes were dissociated in the solutions. Their justifications for their choice are as follows.

A5: “Cause I just thought that these two [Option A and B] are wrong 'cause like, K plus, potassium and like iodine are separated. I thought they had to be together.”

A6: “The chlorine ions are not bonded to the K⁺.”

B5: “I am assuming once mixed, they joined to create a compound.” B5 also rejected options A and B by arguing in their written response, “Because at the end iodine and K and Cl are all still separate.”

B7: “I thought potassium iodide was a compound, and A and B showed ions individually rather than a compound.”

We also observed a similar line of justification in Reaction 2, where students A5 and A6 argued in the written response to reject option B, as A5 wrote, “This is wrong because SO₄ and Cu²⁺ are separated.” and A6 wrote, “This to me shows that the copper and sulfate are not bonded together.”

Reaction 1 and Reaction 2 included diagram options depicting both bonded and dissociated ions. For several students, this “ion-pair” cognitive resource may have served as the determining factor in their decision to eliminate options that showed dissociated ions (and electron transfer). Hence, these responses reflect a high adherence to the ion-pair cognitive resource.

The data suggest that the ion-pair cognitive resource was likely to have a strong adherence when students determined the best representation for the redox reaction. Yet it did not exist in isolation from other aspects of redox. We identified two ways in which the ion-pair cognitive resource was integral to students’ understanding of the redox reactions.

Coexistence of ion pairs and electron transfer. Having the ion-pair cognitive resource did not mean that students were unable to identify the chemical species that undergo oxidation and reduction. In the reaction of Zn + CuSO₄, many of the students who opted for bonded CuSO₄ indicated there was electron transfer between Cu²⁺ and Zn. For example, A5 and A6 mentioned this in their interview about Reaction 2:

A5: “Because Zinc loses electrons and forms Zn²⁺ ions [dissolved in solution]. Copper(II) ions (Cu²⁺) gain electrons and get reduced to form solid copper (Cu) metal, which deposits on the zinc surface.”

A6: “Electrons are transferred between zinc and copper, then zinc bonds to free sulfates.”

And A6 referenced electron transfer in their written response to Reaction 1:

A6: “Cl oxidation changed from 0 to −1 and I oxidation changed from −1 to 0. Cl also bonds to K⁺.”

These students were also able to identify the different chemical species that underwent oxidation and reduction. As they wrote,

A5: “Zinc went under oxidation and Copper went under reduction.”

A6: “Zinc underwent oxidation (0 to +2) and copper underwent reduction (+2 to 0).”

Similarly, in the reaction KCl + I₂, students A5 and A6 demonstrated their understanding of species undergoing different changes. In their written response, they stated,

A5: “Iodide went under oxidation and chlorine went under reduction.”

A6: “Iodine underwent oxidation (lost electrons to become more positive −1 to 0), and chlorine underwent reduction to become more negative 0 to −1.”

They also referred to the changes with respect to the colour changes during the reactions. In fact, these two students also referred to “electron transfer” (Table 3). However, this was not with reference to the transfer between the iodide and chlorine:

A5: “It has to be potassium and chloride. Potassium is giving electron, the chloride is gaining.”

A6: “The K potassium, I think that's showing the electrons being exchanged… with the chlorine touching the potassium, the K potassium, I think that's showing the electrons being exchanged” (response from interview).

As illustrated by the data, the ion-pair cognitive resource could co-exist with their knowledge of electron transfer between Zn and Cu²⁺. In fact, this was quite a common observation among students (Table 3); the ion-pair cognitive resource did not necessarily hamper students’ understanding of redox reactions as the electron transfer from Zn to Cu²⁺. Nevertheless, the case is more complicated in the reaction involving halogens. While students could identify oxidising and reducing agents, their understanding of the electron transfer might have been guided by the interaction of potassium (which tends to give out electrons in its atomic form) and chlorine (which tends to receive electrons in its atomic form) to form new ion pairs of KCl as a product. In other words, the redox reaction started with having ion pairs, underwent electron transfer, and formed new ion pairs (see Option C for Reaction 1).

Coexistence of ion pairs and switching partners. In addition to the ion-pair cognitive resource, some students also believed that the redox reactions involved “switching partners.” When justifying their choice of Option C for the reaction KI + Cl₂, students A7 and B6 started with the ion-pair cognitive resource, as quoted below:

A7: “It shows that potassium iodide (KI) is a compound as K and I are stuck together. Cl2 is a diatomic molecule which are clearly shown as well as the Cl2 are stuck together.”

B6: “Potassium and Iodide are together in solution.”

They also rejected options A and B, stating,

A7: “It implies that the potassium and iodine are not a compound but ions in the solution. It can be misleading and some might think it's separate ions and not KI as a compound.”

B6: “Potassium and Iodide are not together.”

The responses from students A7 and B6 indicate that the potassium ions and iodide ions being in ion pairs was the key factor in determining the best representation for the redox reaction (“Potassium and Iodide are/are not together”). This provides another piece of evidence for ion pairs being a strongly adhered-to cognitive resource among some students. Student A7 associated the concept of compounds with ion pairs, in that KI is a compound only when the ions stay together. This suggests that the ion-pair cognitive resource does not exist in isolation but is associated with other key chemical concepts.

Compared with the students in the previous section who referred to “electron transfer,” these two students did not mention the process and described the reactions of KI + Cl₂ as a simple spatial rearrangement of particles similar to the most generic notion of the displacement reaction AB + C → AC + B. Students A7 and B6 explained the redox reaction as follows:

A7: “During the reaction it shows that Cl₂ is reacting with the K part of the Kl.”

B6: “Elements switch bonding.”

Similarly, in the reaction Zn + CuSO₄, they reiterated spatial switching when writing about why they chose option C:

A7: “It shows CuSO₄ as a whole molecule in itself and shows specifically how the switch between Cu and Zn happens.”

B6: “Zinc and copper switch places.”

It is worth noting that the switching-partner cognitive resource was observed in both reactions (or in different contexts), suggesting that it could be a prevalent cognitive resource. In fact, we noted a degree of coherence between the ion-pair cognitive resource and the switching-partners cognitive resource. That is, a reaction initially involves some ion pairs as reactants, whereby the pair detach from each other and then one of the components attaches to another reactant to form new ion pairs as the products. Modelling students’ reasoning this way helps to explain why student A7 rejected option B for Zn + CuSO₄:

A7: “It doesn’t accurately show which compounds are formed after the reaction nor what compounds are present in the solution, only the ions. It may be a bit confusing.”

Student A7's responses indicate that they believed proper representations of the reaction should involve distinctive changes in ion pairs as compounds before and after the reaction. Thus, it is likely that the student strongly adhered to the ion-pair and switching-partner cognitive resources when making sense of the reaction Zn + CuSO₄. Also, it is apparent that these cognitive resources are consistent with the symbolic representations Zn + CuSO₄ → Cu + ZnSO₄, which may inform or reinforce the student's ion-pair and switching–partner cognitive resources.

Lastly, although these two students did not refer to electron transfer and might have adhered to the switching-partner cognitive resource, they recognised both Zn + CuSO₄ and KI + Cl₂ as redox reactions and were able to tell which was oxidised and which was reduced:

A7 (KI + Cl₂): “Iodine got oxidised, chlorine got reduced.”

B6 (Zn + CuSO₄): “Zinc was oxidised and copper was reduced.”

In closing this subsection, we would like to reiterate that the ion-pair cognitive resource (as indicated by different students in both reactions) is likely to be a strongly adhered-to and prevalent cognitive resource. Also, it could coexist with the switching-partner cognitive resource. Yet, holding ion pair cognitive resource in the contexts of these two reactions does not rule out the possibility of students recognising which chemical species were oxidising agents and which were reducing agents. Table 5 summarises the students who reasoned with ion-pairs while also referring to electron transfer and oxidising/reducing agents, regardless of whether they chose the best diagram in the post-activity questions (Table 2). This suggests that students’ adherence to the ion-pair reflects a priority in reasoning with the cognitive resource rather than a lack of awareness of scientifically relevant ideas.

Table 5 Students who demonstrated using the ion-pair cognitive resource while referring to electron transfer or identifying oxidising and reducing agents

Reactions	Ion pair
	Electron transfer	Oxidising/reducing agents
R1: KI + Cl2	A5, A6, B1,	A5, A6, A7, B1, B5, B6
R2: Zn + CuSO4	A5, A6, B5, B7	A5, A6, A7, B5

---

(2) Ion attraction cognitive resource

Some students tended to believe that redox reactions should involve electrostatic attraction between oppositely charged ions. This is shown by their statements indicating that similarly charged species, such as two negative ions, or elements in the same group (e.g., halogens), did not react. This was particularly evident in Reaction 1 (KI + Cl₂). For example, students B2 and B3 rejected Option A, stating in their written responses,

B2: “I don’t think I⁻ and Cl would react together more readily than K⁺ and Cl because I and Cl are both halogens.”

B3: “Both are negative ions, so it doesn’t work.”

Similarly, B3 and B4 indicated in their written response,

B3: “[The reason for choosing B is] because K is a positive ion and Cl would be a negative ion.”

B4: “[The reason for choosing B is because] ions with right matches,” and they rejected A because “Cl is an anion.”

In the context of KI + Cl₂, students who held this cognitive resource of “ion attraction” often identified potassium and chloride as the reacting species, while iodide (I⁻) was described as a spectator. This cognitive resource is consistent with the general electrostatic principle that positive and negative charges attract, though it was not the driving force for the chemical reaction between KI + Cl₂. The “ion-attraction” cognitive resource was found to coexist with “electron transfer.” For example,

B2: “K⁺ gained electrons so it was reduced and Cl donated/lost electrons so was oxidized” (written response).

The data suggest that even students who were aware of redox reactions involving electron transfer could make sense of the reactions in terms of “ion attraction.”

Although the “ion-pair” and “ion-attraction” cognitive resources share similarities, they have subtle differences. While the ion-pair cognitive resource is about ions remaining paired or bonded, the ion-attraction cognitive resource focuses on the attraction between positive and negative entities during a reaction. For instance, while students B3 and B4 selected Option B (which showed dissociated ions), meaning that they did not seem to reason with the ion-pair cognitive resource, their reasoning was guided by their knowledge of the electrostatic principles that are a driving force for reactions, and they focused on the idea that reactions involved oppositely charged ions.

Compared with the ion-pair cognitive resource that we could infer from students’ understanding of both reactions, the ion-attraction cognitive resource was limited to the context of KI + Cl₂ (Reaction 1). This pattern appears not merely because the reaction 1 provided varying options for which species were reacting, but it was also due to the absence of contextual triggers for the ion-attraction cognitive resource in the context of CuSO₄ + Zn (Reaction 2). In the case of Reaction 1, the presence of varying reacted species (I⁻ and Cl₂vs. K⁺ and Cl₂ options) prompted students with the ion-attraction cognitive resource to reject scientifically accurate options in the belief that only oppositely charged ions would react. However, in Reaction 2 (Zn + CuSO₄), none of the students explicitly referenced “ion attraction” as a driving force of the reaction when explaining their choices. This suggests that the ion-attraction cognitive resource was not universally activated but was instead dependent on the reactions themselves. In other words, the ion-attraction cognitive resource was not prevalent in different contexts.

A further pattern observed in the data was the impact of reaction contexts on students’ reasoning. Students tended to perform better in the metal displacement reaction (Zn + CuSO₄) than in the halogen displacement reaction (KI + Cl₂). A possible reason is that the students were more familiar with reactions involving metals, which are commonly emphasised in introductory chemistry education due to their observable change (e.g., in metal reactivity series) (Haigh et al., 2011; Ortiz Nieves et al., 2012; Wang et al., 2022). In addition, in this study, our animation activity also showed a redox reaction between copper and AgNO₃ (metal displacement) and participants prior lectures used metal reaction to learn about balancing redox reactions and the half-equations which makes the metal reaction more familiar to them.

Another reason it may be more direct to mentally visualise metal reactions. At the submicroscopic level, a metal is represented by a cluster of circles (representing many atoms), and metal ions are shown floating in solution. These features help students follow a simple redox mechanism: Metal A (solid) gives electrons to Metal B's ions, with atoms of Metal A become ions, ions of Metal B become atoms/a solid. In contrast, in a halogen reaction (e.g., as in reaction 1), all the particles, including spectator ions, are shown as small individual or paired circles with charges (positive or negative). This may not be immediately clear to readers which particles interact with which, and may trigger the ion-attraction cognitive resource. The challenges were shown by their responses in which they might have assumed K⁺ and Cl₂ were reacting because ‘positive’ and ‘negative’ attract. Also, students may find it difficult to conceive that I⁻ and Cl₂ could react because they are both halogens, which are negatively charged in their ions.

Discussion

This study investigated how undergraduate students made sense of redox reactions at the submicroscopic level after engaging with animation-based learning activities. Guided by the KiP (diSessa, 1993; Hammer, 1996) and adherence/prevalence view of conceptual learning (Potvin and Cyr, 2017), we examined students’ choices, written responses, and interview data relating to the post-activity questions. Through analysing the data, we identified those cognitive resources that students might have used and those that they strongly adhered to. To illustrate these patterns of adherence, we can return to Fig. 1. The consistent activation of the pentagons could be regarded as the high adherence of ion-pair cognitive resource, whereas other activated cognitive resources, such as electron transfer or oxidising/reducing agents (represented by other bold shapes), played a less determining role. The following subsections discuss the meanings of the findings and how they may contribute to the existing literature on students’ learning in chemistry.

(1) Ion-pair cognitive resource

In the contexts of redox reactions and solution chemistry, it has been consistently shown that many students believe that, when electrolytes are dissolved in water, ions remain bonded in solution (Rosenthal and Sanger, 2012; Kelly et al., 2017; Hansen et al., 2019). Data from this study also demonstrated a similar phenomenon. More importantly, we found that, for some students, their belief that “they bonded together” was key to their determination of the best visual representations for the redox reactions CuSO₄ + Zn and KI + Cl₂. It also played a key role when they were ruling out options that they believed were inaccurate because “they are separated.” This occurred despite the animations they watched, which included verbal statements and visual representations indicating that (1) ions of electrolytes dissociate when they are dissolved in water, and (2) there is a transfer of electrons across reacting species. The results suggest that watching the multimodal animation of ion dissociation had little observable impact on students’ responses in this task. Also, nearly all students in their written responses and/or interviews indicated that the reactions CuSO₄ + Zn and KI + Cl₂ involved “electron transfer” (as in Table 3), meaning that it is likely they had some cognitive resources about electron transfer in these redox reactions. Nevertheless, it was the “ion pairs” (not electron transfer) that informed their decision about the best representations of redox reactions. In this sense, these students had a strong adherence to the ion-pair cognitive resource.

Our data also showed that the ion-pair cognitive resource coexisted with and was related to “electron transfer” across species (e.g., CuSO₄ and Zn; Cl₂ and I⁻; and within K and Cl in KCl). In some other cases, the ion-pair cognitive resource coexisted with the simple spatial rearrangement of chemical species without making reference to electron transfer (e.g., from CuSO₄ + Zn to Cu + ZnSO₄; from KI + Cl₂ to KI + Cl₂). While we only examined students’ understanding of redox in two contexts, the ion-pair cognitive resource was common across both reactions: seven out of 15 students expressed this perspective with regard to reaction 1, and five out of the 11 students who responded to reaction 2 also demonstrated this reasoning (whether or not it involved electron transfer). This finding suggests that the ion-pair cognitive resource could be prevalent in different contexts and among students who had diverse understandings of redox.

One may argue that the ion-pair cognitive resource was activated in this study because both animations and post-animation questions did not represent water molecules and hydrated ions, which made direct pairing of ions a more accessible option for students. Although this simplification aimed to reduce complexity and to make it more direct for students to map the chemical species in the animations with the corresponding ionic equations, we also believe that there are values in representing hydration and hydrated ions. For example, VisChem animations (Tasker and Dalton, 2006) explicitly show hydrated ions and their surrounding water molecules. Studies using this approach in dissolution contexts show that such hydration cues could help students recognise that ions exist as dispersed/solvated species rather than bonded pairs (Magnone and Yezierski, 2024a, 2024b). However, Rosenthal and Sanger (2012) found that half of the students (N = 55) misidentified the water molecules shown in the VisChem redox animation as nitrate ions. In other words, the value of representing hydrated ions in animations seems to be inconclusive. On the one hand, they can support learning and prevent students from reasoning with ion-pair. On the other hand, some students interpreted the water molecules as spectator ions (i.e., NO₃⁻). Yet, these studies supported the notion that cognitive resources are sensitive to contexts. It is an empirical question of how to support students’ fruitful use of the cognitive resources (as in Talanquer, 2024). That is, to what extent including an additional process of metal ions hydration weakens students’ adherence to ion-pair and develops their adherence to electron transfer.

Beyond the context of this study, we suggest that the ion-pair cognitive resource could be related to or even an integral part of the way that students make sense of other fundamental chemical concepts. Drawing on the existing literature, we suggest that several contributing factors may explain the strong adherence and prevalence of ion-pair.

First, symbolic representations provide perceptual cues for students to make sense of submicro phenomena (Wu et al., 2001; Taber, 2009). For example, ionic formulae such as CuSO₄(aq) and KCl(aq) may suggest to students that the ion pairs exist together in aqueous solutions. One of the students in this study indicated that ions had to be together in order to be “compounds”. Upon reflection, we noticed that compounds are often defined in terms of more than one element chemically bonded together. Thus, having dissociated ions in solutions may seem to violate such a common definition and is not consistent with the way that ions are represented in ionic formulae such as “CuSO₄” and “KCl.”

Second, outside the context of redox, it has been shown consistently that some students tend to visualise ion pairs in electrolytes in water (Smith and Metz, 1996), in the solid state of ionic structures (Taber, 1997), and even in lattices (Vladušić et al., 2016). Also, the ion-pair cognitive resource could be regarded as cohering with the simple spatial rearrangement understanding of reactions (e.g., AB + C → AC + B), which is akin to Lego® block rearrangement. According to some students, the simple rearrangement did not have to involve electron transfer (Cheng and Gilbert, 2017). These students understood the ion pair (e.g., A–B) as the initial state of reactions, after which bond breaking occurs (e.g., A and B) and then bond formation (e.g., A–C) (Cheng and Gilbert, 2017; Nehring and Schanze, 2025). The findings of this study may seem to raise a challenge to the value of the simple spatial rearrangement model of reactions as conceived by Cheng and Gilbert (2017). Nevertheless, we are of the view that ion-pair has values, for example, in making sense of precipitation reactions and acid–base reactions (in terms of H⁺ + OH⁻ → H₂O). However, we must be cautious about the limitations of ion-pair when teaching about redox reactions. Such a stance is consistent with the contextual nature of using cognitive resources.

Third, and building on the previous point, in the context of redox, Brandriet and Bretz (2014) found that 70% of the sample (∼1500 first-year university students taking an introductory chemistry course) believed that, in the reaction CdSO₄(aq) + Fe(s) → FeSO₄(aq) + Cd(s), there was electron transfer during the bond breaking of cadmium and sulphate and the bond formation between the iron and sulphate. Most strikingly, the average confidence of different clusters among the participants was 55–60%. The findings, along with another study that identified students’ understanding of MgO as ion pairs formation in the reaction between magnesium oxide (Cheng, 2018), lend support to our finding that students could have a high adherence to the ion-pair cognitive resource, and this cognitive resource could be associated with their understanding of the electron transfer process in redox reactions.

To sum up, the question of whether ions exist in pairs or in a dissociated form may not appear to be directly related to students’ understandings of redox in terms of electron transfer. Previous studies have reported that some students tend to regard ions as existing in pairs (Kelly et al., 2021). Nonetheless, after interpreting our data in light of conceptual adherence, we would like to suggest that students used the ion-pair cognitive resource to explain phenomena, and they prioritised this cognitive resource over others to inform their decisions. More importantly, the ion-pair cognitive resource could be integral to students’ understandings of redox reactions. Some students relied on them more heavily than on the process of electron transfer, and they may have continued to use it even when they correctly described electron transfer. Drawing on studies that investigated students’ understanding of other chemical ideas—such as chemical formulae, chemical reactions, chemical bonding, and solution chemistry—it emerged that the ion pair seems to be a significant cognitive resource that students may make use of in their sense-making of chemistry more broadly.

From a KiP perspective (diSessa, 1993), students’ reasoning that ions “stay together” can be viewed as a p-prim derived from everyday and symbolic experiences, such as perceiving compounds as fixed entities (e.g., NaCl being written as a unit). This intuitive knowledge element, often tacitly reinforced by formulaic representations, may be easily triggered when students view ionic species in solution. When confronted with diagrams that varied in (1) ion distribution, (2) the species involved in electron transfer, and (3) the representation of charge, students showed strong sensitivity to the first feature—how ions were spatially arranged. This suggests that their reasoning was primarily driven by the visual cue of proximity rather than a mechanistic understanding of electron transfer.

(2) Ion-attraction cognitive resource

In this study, some students demonstrated adherence to the ion-attraction cognitive resource to make sense of redox reactions at the submicroscopic level. Particularly, they stated that K⁺ and Cl⁻ should react because they are oppositely charged, and they rejected the option that showed the “same charges” (e.g., Cl and I⁻) because they “could not react.” In fact, ion attraction can be regarded as an electrostatic attraction between oppositely charged species, which is a foundational principle in chemistry and forms a basis for chemical bonding, lattice formation, and solubility. The widely used electrostatic principle and the intuitive idea of positive/negative attraction may relate to students’ adherence to this ion-attraction concept when interpreting redox reactions.

The ideas that “opposite charges attract” and “like charges repel” are relevant in other chemical reactions. A common example is the interaction of positively charged H⁺ and negatively charged OH⁻ in acid–base neutralisation reactions (H⁺ + OH⁻ → H₂O), which students often encounter before learning redox chemistry. This is particularly related to the “ion-attraction” cognitive resource, which is consistent with the idea that the result of attraction between two oppositely charged ions is the neutralisation or cancellation of charges, leading to the formation of a neutral molecule (Boo, 1998). Therefore, it is possible that students’ prior learning makes this ion-attraction cognitive resource a familiar reasoning tool that they may use when trying to make sense of other chemical reactions, including redox processes.

Furthermore, the “ion-attraction” cognitive resource was evident in a previous investigation of students’ understandings of the submicroscopic reaction between Mg and HCl. Cheng and Gilbert (2017) reported that some students regarded the reaction as magnesium atoms losing electrons and becoming positive magnesium ions before reacting with the negatively charged Cl⁻. This idea, that cations and anions must coexist as paired entities due to their opposite charges, has also been previously documented by Zoller (1990) in the context of students’ difficulty with balancing oxidation–reduction equations in a non-neutral aqueous solution. According to this view, no ion exists independently; singular ions imply a process whereby ions of an opposite charge come to maintain electrical neutrality.

From a conceptual learning perspective, this pattern can be interpreted as the application of a scientific idea in an inappropriate context (diSessa, 1993). The KiP perspective would classify “opposites attract” as a p-prim: an intuitive, context-sensitive knowledge element that is easily triggered by surface cues (e.g., seeing a cation and an anion together). When this p-prim is activated in redox contexts, it could interfere with deeper mechanistic reasoning. In our study, some students may not have merely activated this resource momentarily but also relied on it consistently to reject scientifically accurate representations. The data suggest high adherence to “electron attraction” as opposed to “electron transfer” as a key process for redox.

The ion attraction cognitive resource emerged most prominently in in students’ responses to questions about the halogen displacement reaction (KI + Cl₂) but not in their responses to the metal displacement reaction (CuSO₄ + Zn). This may reflect differences in student familiarity: metal reactions are often taught early and with clear visual cues (e.g., solid forming), while halogen reactions may be less emphasised. The results suggest that the ion-attraction cognitive resource could be strongly adhered to (in students’ understandings of KI + Cl₂) but was not as prevalent as the ion-pair cognitive resource, which students also drew upon in another context (CuSO₄ + Zn).

Implications for teaching and further research

The findings from this study suggest that students’ reasoning about redox is shaped by their adherence to specific conceptual resources, such as the ion-pair and ion-attraction cognitive resources. These cognitive resources may arise from prior learning (e.g., acid–base reactions), intuitive expectations (e.g., “opposites attract”), and representational cues (e.g., symbolic equations), and they can persist even after students engage with submicroscopic animations. The findings therefore have implications for both animation and instructional design.

The finding that ion-pair and ion-attraction cognitive resources have strong adherence may mean that focusing on the electron transfer in redox reactions is not enough to promote comprehensive conceptual understanding. We may need to address the ion-pair cognitive resource upfront. Before showing students the electron transfer, it may be worth having animation activities that support students’ capacity to visualise how ions behave and are distributed in solutions when electrolytes dissolve in water (e.g., based on Magnone and Yezierski, 2024a, 2024b). This aligns with recent work that has applied the VisChem Approach (e.g., Magnone and Yezierski, 2024b), which combines molecular animations with storyboarding and sequencing to support conceptual understanding. In particular, we could help students understand that, when the ions of a compound dissociate in water, the substance is still a compound (i.e., it changes from a compound in a solid state to the same compound in a solution).

We observed that the ion-attraction cognitive resource had a stronger adherence when students engaged with the halogen displacement than with the metal reaction. This indicates the potential for progressive sequencing, starting with simple metal displacement reactions to focus on electron transfer then extending to halogen reactions. That is, we could use the metal reaction (or alike) to help students understand that these reactions (1) did not involve positive and negative ion attraction but (2) involved electron transfer. Then, we could encourage students to draw on these two ideas as they learn the redox reactions of halogen displacement. These two types of reactions are therefore not merely examples to illustrate redox. Rather, they can be regarded as progressive examples in which students’ learning of metal redox reactions can be extended to encompass reactions involving halogens.

Given that students’ adherence to ion-pair and ion-attraction emerged within the context of learning about metal and halogen displacement reactions, future research could investigate how these cognitive resources are activated or shift across students’ engagement with a wider range of redox reactions and other contexts, e.g., compound/elements, bonding, dissolution, chemical equations, and chemical formulae. In addition, as this study captured only momentary snapshots of student reasoning, longitudinal work could examine how conceptual adherence develops or transforms over time, particularly as students encounter redox ideas across different contexts or media. Investigating additional influences, such as prior instruction, representational competence, or affective factors (e.g., confidence, interest), may also deepen our understanding of how learners make sense of submicroscopic phenomena.

Finally, although this representational choice was intentional and aimed at reducing complexity, the absence of water molecules may also have shaped how students interpreted ion behaviour, making the ion-pair resource more readily activated. Future research could examine whether including hydrated ions or solvent, such as in VisChem-style animations, alters the cognitive resources students activate when interpreting redox representations.

Limitations

This study has three limitations that we would like to highlight. First, students experienced different instructional conditions during this study, either viewing or critiquing animations. It is possible that these interventions influenced the conceptual resources they used. Although the KiP perspective assumes that context shapes understanding, this study did not collect data on the specific classroom contexts in which students initially viewed the animations. Future research could explore this relationship more systematically to better understand how different types of animation-based instruction affect conceptual adherence and prevalence. Second, the relatively small sample limits the generalisability of the findings to broader student populations. Finally, the study focused on two rather simple redox reactions involving displacement processes, which may have shaped the kinds of cognitive resources students activated; thus, the findings may not fully extend to other redox contexts or chemical systems, but we believe that this is an area worth exploring.

Conclusions

Guided by the Knowledge in Pieces and adherence/prevalence perspectives on conceptual learning, we identified specific ways that students made sense of the structural aspects and processes involved in simple redox reactions. Findings in this study were drawn from data collected through students’ written responses and oral interviews. Students were drawn from two groups who participated in two animation activities; thus, while our sample size was limited, our informants had somewhat diverse learning experiences.

We found that some students strongly adhered to the ion-pair cognitive resource in their sense-making of the two redox reactions. In particular, some students used “ion pairs” as the defining characteristic to determine which animations offered the best representations of redox reactions. Previous studies, both in redox and other concepts in chemistry, have demonstrated that ions remain bonded in solution as a misconception. Based on the Knowledge in Pieces framework, while we identified ion-pair in the topic redox, we suggest that the ion-pair cognitive resource could help make sense of “misconceptions” reported in the literature in diverse topics, e.g., compounds/elements, chemical formulae, solution chemistry, chemical bonding, and structure and properties, as well as other reactions. Moreover, this study found that the adherence of ion-pair could vary in two simple redox reactions. The findings could guide our approach to planning the teaching of redox reactions.

The ion-attraction cognitive resource was found to be strongly adhered to by some students. But it was only evident in one of the reactions (i.e., KI + Cl₂), suggesting that, while this cognitive resource was strongly adhered to, it was not as prevalent as the ion-pair cognitive resource. These two cognitive resources appear to offer some insights into how learning redox was challenging for students, even with animations.

Previous studies have shown that the Knowledge in Pieces perspective can inform students’ learning of chemistry, as well as their learning of physics (Taber and García-Franco, 2010). By integrating this with the adherence/prevalence perspective of conceptual learning (Potvin and Cyr, 2017), we are able to examine students’ learning of chemistry in a new light. Although this study focused on learning with animation, we are confident that the same conceptual framing of this study would be fruitful in other learning contexts.

Author contributions

The first author led the methodology and designed the animations. She did the data collection and analysis. The second author led the conceptual framing and the discussion of the paper.

Conflicts of interest

There are no conflicts to declare.

Data availability

The animation represented in Fig. 2 is available: https://youtu.be/Hg_RwEbSdMQ?si=2XeMDRTkRS4VW_7I.

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information include (1) the link to the animations, (2) details on the animation activities, (3) the differences between each options of the post-activity questions, (4) students responses (written and interview). See DOI: https://doi.org/10.1039/d5rp00192g.

Acknowledgements

The authors gratefully acknowledge the voluntary participation of the students involved in this study. We are sincerely grateful to Dr Ruth Cink and Dr Courtney Ruha for their valuable support, including their agreement to participate, their guidance on the animation design, and their feedback on our initial findings. We also acknowledge the School of Chemistry for facilitating data collection for this research. This study was supported by funding from the Indonesia Endowment Fund for Education (LPDP) as part of the first author's scholarship programme.

References

1. Akaygun S. and Adadan E., (2019), Revisiting the understanding of redox reactions through critiquing animations in variance, in M. Schultz, S. Schmid and G. A. Lawrie (ed.), Research and Practice in Chemistry Education. Singapore: Springer, pp. 7–29 DOI:10.1007/978-981-13-6998-8.
2. Alqadri Z. and Munawwarah M., (2025), Mental models and conceptual change in chemistry: a literature-based perspective on learning challenges, J. Educ. Sci., 9(2), 811–826 DOI:10.31258/jes.9.2.p.811-826.
3. Boo H. K., (1998), Students’ understandings of chemical bonds and the energetics of chemical reactions, J. Res. Sci. Teach., 35(5), 569–581. DOI:10.1002/(SICI)1098-2736(199805)35:5%3C569::AID-TEA6%3E3.0.CO;2-N.
4. Brandriet A. R. and Bretz S. L., (2014), Measuring meta-ignorance through the lens of confidence: Examining students’ redox misconceptions about oxidation numbers, charge, and electron transfer, Chem. Educ. Res. Pract., 15(4), 729–746 10.1039/C4RP00129J.
5. Cheng M. M. W., (2018), Students’ visualisation of chemical reactions—insights into the particle model and the atomic model, Chem. Educ. Res. Pract., 19(1), 227–239 10.1039/C6RP00235H.
6. Cheng M. M. W. and Gilbert J. K., (2017), Modelling students’ visualisation of chemical reaction, Int. J. Sci. Educ., 39(9), 1173–1193 DOI:10.1080/09500693.2017.1319989.
7. Cheng M. M. W. and Leung J. S. C., (2021), Critical thinking across disciplines in university general education: Obesity as a socioscientific issue, in Berry A., Buntting C., Corrigan D., Gunstone R. and Jones A. (ed.), Education in the 21st Century: STEM, Creativity and Critical Thinking, Springer International Publishing, pp. 119–135 DOI:10.1007/978-3-030-85300-6_8.
8. De Jong O. and Treagust D. F., (2003), The teaching and learning of electrochemistry, in Gilbert J. K., Jong O., Justi R., Treagust D. F. and Driel J. H. (ed.), Chemical Education: Towards Research-based Practice, Dordrecht: Springer, pp. 317–337 DOI:10.1007/0-306-47977-X.
9. diSessa A. A., (1993), Toward an epistemology of physics, Cogn. Instr., 10(2–3), 105–225 DOI:10.1080/07370008.1985.9649008.
10. diSessa A. A., (2017), Conceptual change in a microcosm: Comparative learning analysis of a learning event, Hum. Dev., 60(1), 1–37.
11. Gilbert J. K. and Treagust D., (2009), Multiple Representations in Chemical Education. Dordrecht: Springer DOI:10.1007/978-1-4020-8872-8.
12. Hadinugrahaningsih T., Rahmawati Y. and Suryani E., (2022), An analysis of preservice chemistry teachers’ misconception of reduction–oxidation reaction concepts, J. Technol. Sci. Educ., 12(2), 448–465 DOI:10.3926/jotse.1566.
13. Haigh M., France B. and Gounder R., (2011), Compounding confusion? When illustrative practical work falls short of its purpose—A case study, Res. Sci. Educ., 42(5), 967–984 DOI:10.1007/s11165-011-9226-5.
14. Hammer D., (1996), Misconceptions resources or p-prims: How may alternative perspectives of cognitive structure influence instructional perceptions and intentions? J. Learn. Sci., 5(2), 97–127.
15. Hammer D., (2000), Student resources for learning introductory physics, Am. J. Phys., 68, 52–59 DOI:10.1119/1.19520.
16. Hammer D., Elby A., Scherr R. E. and Redish E. F., (2005), Resources, framing, and transfer, in Mestre J. P. (ed.), Transfer of Learning from a Modern Multidisciplinary Perspective, Greenwich, CT: Information Age Publishing, pp. 89–119.
17. Hansen R., Hu B., Riedlova D., Kelly R. M., Akaygun S. and Villalta-Cerdas A., (2019), Critical consumption of chemistry visuals: Eye tracking structured variation and visual feedback of redox and precipitation reactions, Chem. Educ. Res. Pract., 20(4), 837–850 10.1039/C9RP00015A.
18. Kelly R. M., (2016), Exploring the instructional use of contrasting molecular animations of a redox reaction, in Schultz M., Schmid S. and Holme T. (ed.), Technology and Assessment Strategies for Improving Student Learning in Chemistry. Washington, DC: American Chemical Society, pp. 117–136.
19. Kelly R. M., (2017), Learning from contrasting molecular animations with a metacognitive monitor activity, Educ. Química, 28(3), 181–194 DOI:10.1016/j.eq.2017.02.003.
20. Kelly R. M., Akaygun S., Hansen S. J. R. and Villalta-Cerdas A., (2017), The effect that comparing molecular animations of varying accuracy has on students’ submicroscopic explanations, Chem. Educ. Res. Pract., 18(4), 582–600 10.1039/C6RP00240D.
21. Kelly R. M., Akaygun S., Hansen S. J. R., Villalta-Cerdas A. and Adam J., (2021), Examining learning of atomic level ideas about precipitation reactions with a resources framework, Chem. Educ. Res. Pract., 22, 886–904 10.1039/D0RP00071J.
22. Kiernan N. A., Manches A. and Seery M. K., (2024), Resources for reasoning of chemistry concepts: Multimodal molecular geometry, Chem. Educ. Res. Pract., 25(2), 524–543 10.1039/D3RP00186E.
23. Leung J. S. C. and Cheng M. M. W., (2020), Conceptual change in socioscientific issues: Learning about obesity, Int. J. Sci. Educ., 42(18), 3134–3158 DOI:10.1080/09500693.2020.1856966.
24. Magnone K. Q. and Yezierski E. J., (2024a), Applying the VisChem approach in high school classrooms: Chemical learning outcomes and limitations, J. Chem. Educ., 101(3), 727–740 DOI:10.1021/acs.jchemed.3c00827.
25. Magnone K. Q. and Yezierski E. J., (2024b), Generating an evidence-based guide to scaffolding sodium chloride dissolution using the VisChem approach, J. Chem. Educ., 101(4), 1416–1424 DOI:10.1021/acs.jchemed.3c00989.
26. Masykuri M., Afifa F. N. and Ashadi, (2019), Students’ misconceptions on basic concept of redox reaction, 2nd International Conference on Science, Mathematics, Environment, and Education, 2194(1), 020062 DOI:10.1063/1.5139794.
27. Nehring A. and Schanze S., (2025), Turning the plurality of chemistry into a resource for learning: a core competency of chemistry teachers, Sci. Educ., 28, 2051–2078 DOI:10.1007/s11191-025-00624-5.
28. Ortiz Nieves E. L., Barreto R. and Medina Z., (2012), JCE classroom activity #111: redox reactions in three Representations. J. Chem. Ed., 89(5), 643–645 DOI:10.1021/ed100694m.
29. Österlund L.-L., Berg A. and Ekborg M., (2010), Redox models in chemistry textbooks for the upper secondary school: Friend or foe? Chem. Educ. Res. Pract., 11(3), 182–192 10.1039/C005467B.
30. Pölloth B., Diekemper D. and Schwarzer S., (2023), What resources do high school students activate to link energetic and structural changes in chemical reactions? A qualitative study, Chem. Educ. Res. Pract., 24(4), 1153–1173 10.1039/D3RP00068K.
31. Potvin P., (2017), The coexistence claim and its possible implications for success in teaching for conceptual “change”, Eur. J. Sci. Math. Educ., 5(1), 55–66 DOI:10.30935/scimath/9497.
32. Potvin P. and Cyr G., (2017), Toward a durable prevalence of scientific cognitive resources: Tracking the effects of two interfering misconceptions about buoyancy from preschoolers to science teachers, J. Res. Sci. Teach., 54(9), 1121–1142 DOI:10.1002/tea.21396.
33. Rodriguez J.-M. G., Stricker A. R. and Becker N. M., (2020), Students’ interpretation and use of graphical representations: insights afforded by modeling the varied population schema as a coordination class, Chem. Educ. Res. Pract., 21(2), 536–560 10.1039/C9RP00249A.
34. Rosenthal D. P. and Sanger M. J., (2012), Student misinterpretations and misconceptions based on their explanations of two computer animations of varying complexity depicting the same oxidation–reduction reaction, Chem. Educ. Res. Pract., 13(4), 471–483 10.1039/C2RP20048A.
35. Sevian H. and Talanquer V., (2014), Rethinking chemistry: a learning progression on chemical thinking, Chem. Educ. Res. Pract., 15(1), 10–23 10.1039/C3RP00111C.
36. Sherin B. L., Krakowski M. and Lee V. R., (2012), Some assembly required: How scientific explanations are constructed during clinical interviews, J. Res. Sci. Teach., 49(2), 166–198 DOI:10.1002/tea.20455.
37. Shtulman A., (2025), Conflicting views of nature and their impact on evolution understanding, Sci. Educ., 34(4), 1913–1932 DOI:10.1007/s11191-024-00568-2.
38. Shtulman A. and Valcarcel J., (2012), Scientific knowledge suppresses but does not supplant earlier intuitions, Cognition, 124(2), 209–215 DOI:10.1016/j.cognition.2012.04.005.
39. Shtulman A. and Young A. G., (2024), Tempering the tension between science and intuition, Cognition, 243, 105680 DOI:10.1016/j.cognition.2023.105680.
40. Skelling-Desmeules Y., Brault Foisy L.-M., Potvin P., Lapierre H. G., Ahr E., Léger P.-M., Masson S. and Charland P., (2021), Persistence of the “Moving Things Are Alive” Heuristic into Adulthood: Evidence from EEG. CBE—Life Sci. Educ., 20(3), ar45 DOI:10.1187/cbe.19-11-0244.
41. Smith III J. P., diSessa A. A. and Roschelle J., (1994), Misconceptions reconceived: a constructivist analysis of knowledge in transition, J. Learn. Sci., 3(2), 115–163 DOI:10.1207/s15327809jls0302_1.
42. Smith K. J. and Metz P. A., (1996), Evaluating student understanding of solution chemistry through microscopic representations, J. Chem. Ed., 73(3), 233–235 DOI:10.1021/ed073p233.
43. Taber K. S., (1997), Student understanding of ionic bonding: Molecular versus electrostatic framework? Sch. Sci. Rev., 78(285), 85–95.
44. Taber K. S., (2009), Learning at symbolic level, in Gilbert J. K. and Treagust D. (ed.), Multiple Representations in Chemical Education, Dordrecht: Springer, pp. 75–105 DOI:10.1007/978-1-4020-8872-8_5.
45. Taber K. S., (2013), Revisiting the chemistry triplet: Drawing upon the nature of chemical knowledge and the psychology of learning to inform chemistry education, Chem. Educ. Res. Pract., 14(2), 156–168 10.1039/C3RP00012E.
46. Taber K. S. and García-Franco A., (2010), Learning processes in chemistry: drawing upon cognitive resources to learn about the particulate structure of matter, J. Learn. Sci., 19(1), 99–142 DOI:10.1080/10508400903452868.
47. Talanquer V., (2024), How we simplify, circumvent, or distort causal mechanistic reasoning in chemistry, J. Chem. Educ., 101(5), 1793–1797 DOI:10.1021/acs.jchemed.4c00281.
48. Tasker R. and Dalton R., (2006), Research into practice: Visualisation of the molecular world using animations, Chem. Educ. Res. Pract., 7(2), 141–159 10.1039/B5RP90020D.
49. Teichert M. A., Tien L. T., Anthony S. and Rickey D., (2008), Effects of context on students’ molecular-level ideas, Int. J. Sci. Ed., 30(8), 1095–1114 DOI:10.1080/09500690701355301.
50. Vladušić R., Bucat R. B. and Ožić M., (2016), Understanding ionic bonding—a scan across the Croatian education system, Chem. Educ. Res. Pract., 17(4), 685–699 10.1039/C6RP00040A.
51. Wang L., Hodges G. W. and Lee J., (2022), Connecting macroscopic, molecular, and symbolic representations with immersive technologies in high school chemistry: the case of redox reactions. Educ. Sci., 12(7), 428–428 DOI:10.3390/educsci12070428.
52. Wu H.-K., Krajcik J. S. and Soloway E., (2001), Promoting understanding of chemical representations: Students’ use of a visualization tool in the classroom, J. Res. Sci. Teach., 38(7), 821–842 DOI:10.1002/tea.1033.
53. Zaimi I., Watts F. M., Kranz D., Graulich N. and Shultz G. V., (2025), “That's not a super important point”: second-semester organic chemistry students’ lines of reasoning when comparing substitution reactions, Chem. Educ. Res. Pract., 26(1), 112–125 10.1039/D4RP00086B.
54. Zoller U., (1990), Students’ misunderstandings and misconceptions in college freshman chemistry (general and organic), J. Res. Sci. Teach., 27(10), 1053–1065 DOI:10.1002/tea.3660271011.

---